-
Notifications
You must be signed in to change notification settings - Fork 1.9k
/
max_pool_with_argmax.ts
79 lines (72 loc) · 3.6 KB
/
max_pool_with_argmax.ts
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
/**
* @license
* Copyright 2018 Google LLC. All Rights Reserved.
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
* =============================================================================
*/
import {ENGINE} from '../engine';
import {MaxPoolWithArgmax, MaxPoolWithArgmaxAttrs, MaxPoolWithArgmaxInputs} from '../kernel_names';
import {NamedAttrMap} from '../kernel_registry';
import {Tensor, Tensor4D} from '../tensor';
import {NamedTensorMap} from '../tensor_types';
import {convertToTensor} from '../tensor_util_env';
import {TensorLike} from '../types';
import {op} from './operation';
/**
* Computes the 2D max pooling of an image with Argmax index.
* The indices in argmax are flattened, so that a maximum value at position `[b,
* y, x, c]` becomes flattened index: `(y * width + x) * channels + c` if
* include_batch_in_index is False; `((b * height + y) * width + x) * channels
* +c` if include_batch_in_index is True.
*
* The indices returned are always in `[0, height) x [0, width)` before
* flattening.
*
* @param x The input tensor, of rank 4 or rank 3 of shape
* `[batch, height, width, inChannels]`. If rank 3, batch of 1 is assumed.
* @param filterSize The filter size: `[filterHeight, filterWidth]`. If
* `filterSize` is a single number, then `filterHeight == filterWidth`.
* @param strides The strides of the pooling: `[strideHeight, strideWidth]`. If
* `strides` is a single number, then `strideHeight == strideWidth`.
* @param dataFormat An optional string from: "NDHWC", "NCDHW". Defaults to
* "NDHWC". Specify the data format of the input and output data. With the
* default format "NDHWC", the data is stored in the order of: [batch,
* depth, height, width, channels]. Only "NDHWC" is currently supported.
* @param pad The type of padding algorithm.
* - `same` and stride 1: output will be of same size as input,
* regardless of filter size.
* - `valid`: output will be smaller than input if filter is larger
* than 1x1.
* - For more info, see this guide:
* [https://www.tensorflow.org/api_docs/python/tf/nn/convolution](
* https://www.tensorflow.org/api_docs/python/tf/nn/convolution)
* @param includeBatchIndex Defaults to False. Whether to include batch
* dimension in flattened index of argmax.
*
* @doc {heading: 'Operations', subheading: 'Convolution'}
*/
function maxPoolWithArgmax_<T extends Tensor4D>(
x: T|TensorLike, filterSize: [number, number]|number,
strides: [number, number]|number, pad: 'valid'|'same'|number,
includeBatchInIndex = false): NamedTensorMap {
const $x = convertToTensor(x, 'x', 'maxPoolWithArgmax');
const inputs: MaxPoolWithArgmaxInputs = {x: $x};
const attrs:
MaxPoolWithArgmaxAttrs = {filterSize, strides, pad, includeBatchInIndex};
// tslint:disable-next-line: no-unnecessary-type-assertion
const result = ENGINE.runKernel(
MaxPoolWithArgmax, inputs as unknown as NamedTensorMap,
attrs as unknown as NamedAttrMap) as Tensor[];
return {result: result[0], indexes: result[1]};
}
export const maxPoolWithArgmax = /* @__PURE__ */ op({maxPoolWithArgmax_});